Content-Seam-Preserving Multi-Alignment Network for Visual-Sensor-Based Image Stitching

被引:2
|
作者
Fan, Xiaoting [1 ]
Sun, Long [2 ]
Zhang, Zhong [1 ]
Liu, Shuang [1 ]
Durrani, Tariq S. [3 ]
机构
[1] Tianjin Normal Univ, Tianjin Key Lab Wireless Mobile Commun & Power Tra, Tianjin 300387, Peoples R China
[2] Tianjin Univ, Sch Elect & Informat Engn, Tianjin 300072, Peoples R China
[3] Univ Strathclyde, Dept Elect & Elect Engn, Glasgow G1 1XQ, Scotland
关键词
visual-sensor-based image stitching; deep homography; mesh warping; content-preserving; edge-assisted; QUALITY ASSESSMENT; DEEP HOMOGRAPHY; CONSISTENT; CONSTRAINTS; PANORAMAS;
D O I
10.3390/s23177488
中图分类号
O65 [分析化学];
学科分类号
070302 ; 081704 ;
摘要
As an important representation of scenes in virtual reality and augmented reality, image stitching aims to generate a panoramic image with a natural field-of-view by stitching multiple images together, which are captured by different visual sensors. Existing deep-learning-based methods for image stitching only conduct a single deep homography to perform image alignment, which may produce inevitable alignment distortions. To address this issue, we propose a content-seam-preserving multi-alignment network (CSPM-Net) for visual-sensor-based image stitching, which could preserve the image content consistency and avoid seam distortions simultaneously. Firstly, a content-preserving deep homography estimation was designed to pre-align the input image pairs and reduce the content inconsistency. Secondly, an edge-assisted mesh warping was conducted to further align the image pairs, where the edge information is introduced to eliminate seam artifacts. Finally, in order to predict the final stitched image accurately, a content consistency loss was designed to preserve the geometric structure of overlapping regions between image pairs, and a seam smoothness loss is proposed to eliminate the edge distortions of image boundaries. Experimental results demonstrated that the proposed image-stitching method can provide favorable stitching results for visual-sensor-based images and outperform other state-of-the-art methods.
引用
收藏
页数:15
相关论文
共 50 条
  • [21] GMiRec: A Multi-image Visual Recommendation Model Based on a Gated Neural Network
    Mu, Caihong
    Tang, Xin
    Luo, Jiashen
    Liu, Yi
    KNOWLEDGE SCIENCE, ENGINEERING AND MANAGEMENT, PT II, KSEM 2023, 2023, 14118 : 331 - 342
  • [22] Image data visualization and communication system based on sensor network simulation and visual feature extraction
    Xu, Jin
    Measurement: Sensors, 2024, 33
  • [23] Heterogeneous Feature Fusion-based Optimal Face Image Acquisition in Visual Sensor Network
    Lin, Kuicheng
    Wang, Xue
    Cui, Sujin
    Tan, Yuqi
    2015 IEEE INTERNATIONAL INSTRUMENTATION AND MEASUREMENT TECHNOLOGY CONFERENCE (I2MTC), 2015, : 1078 - 1083
  • [24] Visual semantic alignment network based on pre-trained ViT for few-shot image classification
    Zhang, Jiaming
    Wu, Jijie
    Li, Xiaoxu
    2024 ASIA PACIFIC SIGNAL AND INFORMATION PROCESSING ASSOCIATION ANNUAL SUMMIT AND CONFERENCE, APSIPA ASC, 2024,
  • [25] Privacy-preserving multi-source content-based image retrieval for IoT in cloud computing
    Li, Yao
    Wei, Xiaochao
    Cai, Guopeng
    Wang, Hao
    ANNALS OF TELECOMMUNICATIONS, 2024, 79 (7-8) : 553 - 565
  • [26] Deformable registration network based on multi-scale features and cumulative optimization for medical image alignment
    Xing, Lumin
    Liu, Wenjian
    Wang, Xing
    Li, Xin
    Xu, Rui
    Wang, Mini Han
    BIOMEDICAL SIGNAL PROCESSING AND CONTROL, 2024, 95
  • [27] Content-Based Multi-Channel Network Coding Algorithm in the Millimeter-Wave Sensor Network
    Lin, Kai
    Wang, Di
    Hu, Long
    SENSORS, 2016, 16 (07)
  • [28] Fuzzy VA-Files for multi-label image annotation based on visual content of regions
    Ben Rejeb, Imen
    Ouni, Sonia
    Barhoumi, Walid
    Zagrouba, Ezzeddine
    SIGNAL IMAGE AND VIDEO PROCESSING, 2018, 12 (05) : 877 - 884
  • [29] Fuzzy VA-Files for multi-label image annotation based on visual content of regions
    Imen Ben Rejeb
    Sonia Ouni
    Walid Barhoumi
    Ezzeddine Zagrouba
    Signal, Image and Video Processing, 2018, 12 : 877 - 884
  • [30] Image-based visual servoing of an omnidirectional mobile robot without velocity sensor using multi-layer artificial neural network dynamics
    Boo J.
    Chwa D.
    Chwa, Dongkyoung (dkchwa@ajou.ac.kr), 1600, Korean Institute of Electrical Engineers (69): : 594 - 601